Modeling Typhoon Event-Induced Landslides Using GIS-Based Logistic Regression: A Case Study of Alishan Forestry Railway, Taiwan

Table 4

Regression coefficients obtained and statistics for the independent variables.

(a)  

Variable		Coefficient	Numerical variables			S.E.	𝑃 value
			Range	Mean	S.D.
Terrain roughness		0.496	12.88∼180.52	59.15	22.30	0.029	<0.01
Profile curvature		0.256	−26.82∼28.21	−0.10	1.74	0.256	<0.01
NDVI		−0.908	−0.46∼0.62	0.31	0.17	0.038	<0.01
NDWI		0.585	−0.59∼0.41	−0.14	0.09	0.032	<0.01
Distance to river		−0.195	0∼5249.76	1260.38	1217.93	0.025	<0.01
Maximum 1-h rainfall intensity (mm/hr)		0.391	76.80~100.53	93.04	5.70	0.391	<0.01
Constant		−0.304

(b)  

Variable	Class	Coefficient	Study area (%)	S.E.	𝑃 value
Aspect	N		13.70
	NE	0.382	11.12	0.080	<0.01
	E	0.403	11.29	0.087	<0.01
	SE	−0.172	10.53	0.098	0.081
	S	−0.35	11.07	0.106	<0.01
	SW	−0.519	13.00	0.114	<0.01
	W	−0.869	13.48	0.106	<0.01
	NW	−0.361	15.83	0.090	<0.01
Lithology	Schist, arenite		10.48
	Sandstone, mudstone	0.031	80.99	0.123	0.013
	Sandstone, schist	0.327	8.53	0.099	<0.01